Developing an Augmented Reality Application

in the Framework of Architecture Degree

Albert Sanchez Riera

Ernest Redondo Dominguez

Dept. of Architectural Graphic

Dept. of Architectural Graphic
Expression II.
Expression I.
UPC (Barcelona Tech)
UPC (Barcelona Tech)
School of Building Construction of
School of Architecture of Barcelona.
Barcelona. EPSEB
ETSAB
C/. Gregorio Maran, 44-50, 08028.
Avda. Diagonal, 649, 2. 08028
Barcelona, Spain
Barcelona, Spain
+34 934016290
+34 934016384

albert.sanchez.riera@upc.edu

David Fonseca Escudero

Dept. of Architecture
La Salle Barcelona
Quatre Camins 2, 08022
Barcelona, Spain
+34 932902430
Universitat Ramon Llull, Spain

fonsi@salle.url.edu

ernesto.redondo@upc.edu

ABSTRACT

1. INTRODUCTION

We describe the development, implementation and evaluation of an

augmented reality (AR) application for mobile devices. It has been
particularly intended to be used in educational environments. This
application has been called U-AR, and is based on optical image
recognition from real environment images, to act as markers. Once
they are associated with virtual information (3D architectural
models, in our case), the application allows viewing, scaling, and
positioning it interactively and in real time. Some improvements
have been implemented compared to existing commercial ones,
such as the ability to view different models consecutively with a
single marker and the possibility to move these computer-generated
objects around the scene.

Specific application development for educational purposes, using

augmented reality technology (AR), is an educational research
project that arises from the Interest Group for Logistics and
Education in Architecture, GILDA. It is an interuniversity group
centered in the architectural field, assigned to the Project RIMA,
(Research and Innovation in learning methodologies) from ICE,
(Institute of Education Sciences at the Polytechnic University of
Catalonia, UPC), which is specialized in the field of education
technological disciplines.
This project aims to respond to the growing need of new
technologies incorporation in learning processes. They are
alternative to traditional ones, and provide new tools to improve
motivation and student academic performance, enabling a greater
acquisition of skills by the student and, as well see, reverting in
better academic performance. To evaluate its feasibility, a case
study was assessed. It was conducted on a Computer Applications
(APF) course at the UPC.

The main objective was to assess the feasibility of using AR on

mobile devices in educational environments. In addition we
evaluated academic performance improvement using this
technology. Validation was done through a case of study. It was
carried out by Architecture students of the ETSAB, (UPCBarcelona-Tech) outdoor. Results obtained by students
architectural proposals, after using this technology, and by
questionnaires responses, demonstrated the application suitability as
a new tool to be used in learning processes.

Which aims to complete student`s training and technical resources

to design, analyze, develop and present architectural projects under
the overall theme of the course of Housing and Urban environment.
This is embodied in two main directions: the utilization of digital
image processing and the use of agile tools to create 3D virtual
scenes. Results demonstrated that better results are achieved when
students use AR technologies on mobile devices (running on
Android and iOs software).

Categories and Subject Descriptors

H.5.1. [Multimedia Information Systems]: Artificial, augmented,
and virtual realities; Evaluation/methodology. K.3.1. [Computer
Uses in Education]: Collaborative learning, Distance learning.

2. THEORETICAL, PEDAGOGICAL AND

DIDACTIC FOUNDATIONS

General Terms

Learning, by definition, is the process by which memories are built,

while
memory
is
the
result
of
learning
[1].
In recent years, the desire of learning process improvement has led a
transition to a technologically enhanced classroom, where
computers, media players, interactive whiteboards, internet, web 2.0
tools, and games have been incorporated. The emergence of the
World Wide Web revolutionized how to access, share and encode
information, playing a key role in learning. E-mail and mobile
phones have transformed the way we communicate, and the list of
technologies that can be useful in learning processes is huge and
constantly growing, not being simple to define which may be
suitable for learning and which are not [2].

Design, Experimentation, Human Factors.

Keywords
Augmented reality, mobile devices, learning processes.

Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that copies
bear this notice and the full citation on the first page. To copy otherwise, or
republish, to post on servers or to redistribute to lists, requires prior specific
permission and/or a fee.
UXeLATE12, November 2, 2012, Nara, Japan.
Copyright 2012 ACM 978-1-4503-1593-7/12/11...$15.00.

More recently, immersive technologies in virtual and augmented

reality worlds have been used. Its usefulness has been assessed by

37

is the foundation of learning, however, students engagement is not

easy..

numerous international projects [3], [4], [5], [6], [7], [8], [9], [10].
These experiences, that use augmented reality in the area of
entertainment and education, demonstrated the great potential of this
technology. But in education, it may still be considered as a new
tool, and further investigation is necessary, paying special attention
to user experience and learning processes [11], because, despite the
ongoing effort to technology implementation, there is also the need
to immerse students in new learning environments, these in turn,
continuously changing [12].

3. U-AR application design.

The application ought to allow the creation of AR content channels
and its visualization via a mobile device. Each channel consists on
3D models series that students have previously generated and
associated to pre-defined markers (limited to five). As we said
before, these markers are images from real world that the
application detects to overlay associated virtual information.

Teachers faces the challenge of constantly be updated to provide

new forms of teaching, focusing on the acquisition of generic skills
in which students must construct their own knowledge through
constructivism, proposed by Piaget [13] and meaningful learning
proposed by Ausubel [14]. Thus, students must relate new
information with that one that already owns, as opposed to
memorize contents through classical exposure and repetition. From
this constructivist approach, learning can be facilitated, but each
person reconstructs his own inner experience. In this way we could
consider that knowledge is unique to each person, and depends of
their own internal reconstruction and subjective reality. The role of
teachers is reduced, in many cases, to provide information to
students, being a guide and facilitator of knowledge acquisition.

One of the advantages that differentiate U-AR other commercial

applications (ArMedia, Aurasma, junaio, layar) is the fact that a
single marker contains several models, which is useful in
architectural learning, to facilitate comparison of different models,
options, or proposals, without varying the reference image. It also
allows the positioning of each object relative to the marker
independently, which is essential to minimize AR optical marker
based recognition systems, that originated so many problems of
stability in the scene [18]. Outdoors, these limitations are: strong
dependence of ambient light conditions, camera distance and marker
size relationship, and the fact that the physical environment is
altered by fiducially markers. Using U-AR, user may be near a
marker, which is readily recognizable, and once the 3D model is
overlaid, it can be moved to the desired position, maintaining the
relationship of size and perspective relative to marker position
(which must be always visible).

In spite of this, we can frequently find situations in the classroom

where educational contents are simply exposed and presented
without any interaction by the student, that receive passively new
concepts to be memorized, getting bored easily and consequently
minimizing their learning. Student motivation is essential to reverse
this situation [10].

For its implementation QCAR Augmented Reality library was used.

It is an augmented reality SDK (software development kit) with
proprietary license developed by Qualcomm Austria Research
Center for Android and IOS platforms. The main difference with
traditional Artoolkit libraries is the use of real images from
environment, instead of typical squared markers.

When students are active and motivated they engage, participate,

and interact with content, cognitive mechanisms of learning are
activated, maximizing the learning. The next question that the
teacher should ask is: Can the technologies of learning (TEL,
Technology Enhanced Learning) change this attitude? For Dror [2],
this change depends on whether technology is used as a means to
promote what he calls the three learning C's: Control, Challenge and
Commitment.

The process has been divided into two parts. On the one hand,
theres a virtual-contents manager program that runs on students
desktop computers, and lets them to create and manage content
channels. User can associate one or several models to one marker
choosing initial visualization scale, and channel and 3d models
name (fig.1). The programming language we decided to use is Java.
One of the biggest advantages of Java is that it runs on a virtual
machine, so that a Java application is independent of the operating
system in which is executed. So it could be used on any computer,
whether Mac, Windows or Linux.

There is where AR technology, through motivation and students

engage, can helps to improve the learning process performance [10],
[15]. Augmented reality and virtual reality (VR) share some
common features such as immersion, navigation and interaction
[16]. However, AR has two main advantages over VR:
Allow collaborative experience in a real scene. So users can work
with computer-generated objects as if they were real objects in a
real environment, in real time; and Tangible Interaction. By
superimposing virtual objects in a real environment through
markers, user can modify and manipulate the scale, position and
location of virtual objects. This interaction "tangible", achieved
simply by modifying these markers, becomes an extremely simple
and natural interface that requires no prior learning by the user. So
we could say that RA technology, by providing new interaction
possibilities, promote student active participation in its own
knowledge construction. Thus, it becomes a suitable medium to be
used in schools [17].
By designing an AR mobile tool we will give students control
ownership of their learning process. Besides that, experience will be
a challenge when student confronted with it, in a new way. Then
learning will be more effective. Finally, we can say that the AR
implementation effort will not help, if student dont make a
commitment effort to complete the exercise. As we see commitment

Fig. 1.- Virtual-content manager interface. A: images used as

markers; B: cannel name; C: initial scale; D: virtual
models selection

38

On the other hand we have the mobile device application, which

recognize markers using the device camera, register them, and
shows virtual information overlapped to the real scene. Application
allows student to alternate easily its own generated content, and also
to move and to scale in an intuitive and tactile interface.

simulation. Because most of the times, the virtual model lacks

realism and are not integrated sufficiently into the scene to be
credible. This problem takes on special significance to visualize
architectural proposals, because a non-integrated scene may seem
unlikely and unattractive (Fig. 5).

Fig. 2.- Logo, installation and application on mobile device.

To solve the problem of lighting conditions differences between real

environment and virtual objects superimposed, students generated
models had to be single meshes, and they incorporated textured
maps simulating lighting conditions from real environment. For
occluders implementation we followed the concept of phantoms
suggested by [19]. The main idea is to generate a model identical to
real-world one, that produces an effect of occlusion on the virtual
object.

When students finished this process, a folder is generated. It

contains an *.xml (Extensible Markup Language) file that has all
channel generated structure, 3dmodels (*.obj and *.mtl) and texture
files, organized in subfolders. This information is needed to be
loaded on application in order to read channel information.

Fig. 5. Sample images from the same 3dmodel, using occluders

and lighting integration modeling techniques.

4. Case Study
In this case, we focused on the study of interventions in the urban
landscape of Barcelona with ICT help. Following these premises,
we worked on the Plaza Flassaders (Fig. 6), who recently
completed the construction of the extension of the Picasso Museum
in 2009.

Fig. 3.- Sample Loading and Displaying generated channel.

Xml file is widely used to describe data structures and allows
communication between different applications. In our case, the
content manager creates an XML file and mobile application
interprets the document to load and display correctly virtual model
(Fig.4).
<Channel>
<name>twoMarkers</name>
<layers>
<Layer>
<markerId>1</markerId>
<name>Marcador 1</name>
<width>20.0</width>
<models>
<Model>
<name>armadura</name>
<filename> model1.obj</filename>
</Model>
<Model>
<name>cadira</name>
<filename> model2.obj</filename>
</Model>
</models>
</Layer>
<Layer>
<markerId>2</markerId>
<name>Marcador 2</name>
<width>20.0</width>
<models>
<Model>
<name>facilities1</name>
<filename> facilities.obj</filename>
</Model>
</models>
</Layer>
</layers>
</Channel>

Fig. 6. Case study location.

The working group is composed of 25 students divided into three
groups: a control group, consisting of 8 users who do not have
smartphones or AR technology is not supported (No_AR). They
will follow the traditional course based on slides; and two
experimental groups, which we call iOs (9 students) and Android (8
students) both will be involved in AR specific training. Android
users will use the U-AR application and iphone users will use
Armedia.

4.1 Research questions

In order to investigate the impact of two technologies
implementation scenarios: S1 (based on slides) and S2 (based on
augmented reality technology on mobile devices), on student
academic performance. The research questions were:

Fig.4.- XML document structure sample.

1. Are there any differences in academic results depending on which

of the two teaching scenarios proposed are used?

For rendering we use OpenGL (Open Graphics Library), as a

standard specification defining a cross-language, multi-platform API
(Application Programming Interface) widely used for writing
applications and simulating physics.

2. Is there any difference in students degree of satisfaction and

motivation depending on which of the two teaching scenarios
proposed they used?

Finally, in order to get a better integration of virtual models in the

scene we were focused on occluders recognition and lighting

39

AR groups used U-AR, they generated their own content channels

and were able to see their proposals in site, overlapped to the real
environment. They also were able to see different models with the
same marker and to verify scale, size, and position comparing
different proposals1.

The study also included a quantitative analysis to determine

usability of U-AR application. Some experiences have been done to
evaluate Virtual Environments (VE) usability.
Stanney [20], used a multi-criteria assessment technique to design
and evaluate Virtual environments. In our case we were based on
ISO 9241-11 which provides usability guidelines [21];

Effectiveness, defined as the users ability to complete tasks

during the course, in relation to the "accuracy and integrity"
that it had been made.
Efficiency, on the assigned resources, they asked questions
related to the expenditure of time and effort for solving the
proposed exercise.
Satisfaction, understood as subjective reactions of users about
the course.

4.2 Procedure

Fig 8. Sample images using U-AR application in site.

The whole course was performed in three phases. Augmented

reality experience was conducted on the second one.

4.2.3 Third phase: Final Project Presentation

In this last phase, both groups were required to reconsider their
proposals. Reformulation of their project was done. Scale and
dimensions of the sculpture, points of view and areas of impact were
changed according with in site information obtained using
traditional an AR scenarios.

4.2.1 First Phase:

In summary, this phase involved the use of all digital architectural
tools knowledge acquired by students throughout their training in
previous courses to model existing environment, to select an artistic
reference, and to generate a virtual scene incorporating their
proposals. Students captured site information through digital
photography and vectorial cartography.
Specialized applications as Photomotch, Ptlens, Sketchup were used
to rectify images, in order to get basic dimensions and
measurements from the environment. They chose an artistic
reference, and modeled their proposals using AutoCAD, Rhino, or
SketchUp. Students models were incorporated to the virtual
generated scene. They were required to define scale and location,
and to render preliminary proposals.

Fig 9. Final presentation results, once students proposals were

reformulated.

5. EVALUATION
To evaluate this exercise all students made an architectural proposal
in a conventional manner, as we said in point 4.2.1 (first phase). In
this case, as mentioned, they designed a sculpture, furniture, or
some kind of intervention for a real space choosing an artistic
reference. After visiting the site, control group and experimental
groups (using AR)
reformulate their proposals. Pre and post
proposals were evaluated using the following criteria:

Fig 7. Preliminary students proposals.

4.2.2 Second Phase:

We called this: Visual analysis and adjustments in site, of their
sculpture and urban design proposals using both traditional
processes and AR.

In this phase we defined experimental and control groups between

the different students of the course. Students were asked about their
availability to participate in the experiment. 8 students continue the
ordinary course (NO_AR) and the other 17 are divided into two
experimental groups. Android (8) and IOS group (9).

30%: Quality of the model presented (textures, light

immersion, no. Of polygons, and no. Of textures used).
40%: Integration into the site (scale, position, viability of the
proposal...)
30%: Subjective opinion about proposal quality (through
other users rating).

Means (from 1to 10) and Std. Deviation of the three groups in the
first phase, previous AR training (PRE), were very similar:

Once they visited the site, control group continued developing their
proposals in traditional way. They used for that 3dsmax Sketchup,
and other 3dmodeling tools. They made metric analysis of the scale,
position and orientation of sculptural models proposed in order to
adjust their proposals.

40

A video can be found at: http://youtu.be/2sJnSv_sgPs

Table 4. Results comparing values obtained in pre-and post-test

with the Student t test for paired series.

Table 1. Previous results of AR training

grup_2

Mean

Std. Deviation

Android

6.1200

1.7154

No_AR

6.1662

2.2129

ios

6.2978

2.0043

Total

25

6.1988

Paired Differences
Mean

1.9057

0.146

Between Groups

Within Groups

87.014 22

Total

87.160 24

PRE--POST_NOAR

Pair 2
Pair 3

(2-

0.09

-2.94

3.12 0.07

0.946

PRE- POST_ios

-1.56

-3.11

-0.19 -2.34

0.048

PRE-POST_And

-1.39

-2.60

-0.17 -2.70

0.031

As we mentioned before, we evaluated users assessment using

questionnaires based on ISO 9241-11, which provides usability
guidelines (effectiveness, efficiency and satisfaction). Responses
average were very similar, ranged from 3.59 to 3.73, out of 5.

Sig.

0.073 0.01

Pair 1

Sig.

5.1 U-AR application usability

Table 2. Anova analysis Between groups.

Mean Square

df

However, in the control group, which carried out the activity

through conventional training, table shows no significant differences
between the mean values i PRE POST, therefore no improvement in
the scores is achieved. It is accepted in this case the null hypothesis
null (Ho).

The analysis of qualifications "PRE" between the three groups

(experimental groups ios and Android, and No_AR, control group),
shows that theres not significant differences between groups
regarding their qualifications before to conduct training (F2, 22 =
0.018, p = 0.982).
df

Lower Upper

To verify that they are equal groups, we used analysis of variance

(ANOVA). The null hypothesis (h0) is that the scores match from
different groups (no differences).

Sum of Squares

95%Conf. Int. dif

0.982

3.955

Upon completion of the proposed course (third phase) students

delivered their proposals to be scored as described above. Results
showed that experimental groups, which participated in the RA
experience, increased their scores.
Table 3. Result of qualifications and gain obtained per group,
with PRE and POST presentations:
grup
Android
No_AR
ios
Total

nota_pre

nota_post

Gain

Mean

6.1200

7.5112

1.3912

Std. Deviation

1.7154

0.9918

1.4559

Mean

6.1662

6.0762

-0.0900

Std. Deviation

2.2128

2.0716

3.6307

Mean

6.2977

7.8611

1.5633

Fig. 10. Final average rating

Std. Deviation

2.0043

1.0982

2.0080

Mean

6.1988

7.1780

0.9792

The overall assessment of the courses was rated 4.18 points out of 5.
This gives an idea of the degree of satisfaction achieved.

Std. Deviation

1.9056

1.5992

2.5245

In a correlation analysis between the course global opinion and the

other variables, a high correlation (0.68) was detected with: the
representativeness of the exercise. So this variable seems crucial to
the success of this kind of teaching experience. Global opinion was
not so correlated with the fact of being able to solve the exercises
independently. The strongest correlation (0.89), however, was with
obtained gain variable. Students with little gain between pre and
post qualifications rated worse than student from experimental
groups (with higher gains). On other hand, variables related to prior
knowledge of technology and to the use of different software and
operating systems did not correlate significantly with the course
global opinion.

To verify that the observed increase, is statistically significant, it is

considered as null hypothesis (Ho), the fact that the mean scores do
not vary at the end of the proposed training, that is, after the AR
training the experimental groups (ios and Android) and the control
group did not increase their results differently.
The obtained values indicate that experimental groups (ios and
Android) have significant differences between mean values
(sig<0,05) before and after the course, thus the null hypothesis (Ho)
is rejected, and we can affirm that these groups had improved
results.

6. Conclusions
In relation to U-AR, and the use of new technology on learning
processes with mobile devices, some advantages have been found
over other commercial applications such as: the possibility to

41

display several 3Dmodels without changing the marker (which

allows to compare different architectural proposals and hypotheses);
and the ability to move objects in the scene. These benefits were
crucial to the viability of the study outdoors, helping to minimize
the problem of stability in the scene of that kind of AR systems,
based on optical recognition.
Related to research questions, we found significant differences in
academic results depending on which of the two teaching scenarios
were used. (Research question 1). And results show significant
differences in students satisfaction and motivation depending on
which of the two teaching scenarios were proposed (research
question 2). However, similar experiences with larger groups must
be repeated to compare these results.

[9]

C.-M. Chen and Y.-N. Tsai, Interactive augmented reality

system for enhancing library instruction in elementary
schools, Computers & Education, vol. 59, no. 2, pp. 638
652, Sep. 2012.

[10]

. Di Serio, M. B. Ibez, and C. D. Kloos, Impact of an

Augmented Reality System on Students Motivation for a
Visual Art Course, Computers & Education, pp. 111,
2012.

[11]

C. Portals Ricart, Live LEGO House: a Mixed Reality

Game for the Edutainment 19 - 28., Schottish Online
Journal of E-Learning, vol. 1, no. 1, pp. 1928, 2007.

[12]

T. N. Arvanitis, A. Petrou, J. F. Knight, S. Savas, S.

Sotiriou, M. Gargalakos, and E. Gialouri, Human factors
and qualitative pedagogical evaluation of a mobile
augmented reality system for science education used by
learners with physical disabilities, Personal and
Ubiquitous Computing, 13(3), pp. 243250, 2007.

Results suggest that combination of an attractive technology and

the user-machine interaction that involves the AR, make students
feel more motivated. They created dynamic photomontage and
architectural proposals, for visual evaluation in a real environment,
getting involved on their own knowledge construction, and
consequently, development and evolution of graphic competences
and space skills were increased in a shorter learning periods, and
their academic performance was significantly improved.

[13]

J. Piaget, H. E. Gruber, and J. J. Vonche, The essential

Piaget. J. Aronson, 1995.

[14]

7. REFERENCES

D. P. Ausubel, Educational psychology: a cognitive view.

Holt, Rinehart and Winston, 1968.

[15]

R. CHEN and X. WANG, An Empirical Study on

Tangible Augmented Reality Learning Space for Design
Skill Transfer, Tsinghua Science & Technology, vol. 13,
no. October, pp. 1318, Oct. 2008.

[16]

M. Dunleavy, C. Dede, and R. Mitchell, Affordances and

Limitations of Immersive Participatory Augmented Reality
Simulations for Teaching and Learning, Journal of
Science Education and Technology, vol. 18, no. 1, pp. 7
22, Sep. 2008.

[17]

C. Portals Ricart, Entornos multimedia de realidad

aumentada en el campo del arte, Universidad politecnica
de Valencia, 2008.

[18]

E. Redondo, A. Snchez Riera, and J. Puig, Gironella

tower in Gerunda, teaching roman architecture, using 3D
modeling and augmented reality. A case study, in
S.A.V.E. Heritage - International Forum S.A.V.E. Heritage
Safeguard of Architectural, Visual, Environmental
Heritage, Capri, 2011, pp.10211029.

[19]

A. Fuhrmann, G. Hesina, F. Faure, and M. Gervautz,

Occlusion in collaborative augmented environments,
Computers & Graphics, vol. 23, no. 6, pp. 809819, Dec.
1999.

[20]

K. M. Stanney, M. Mollaghasemi, L. Reeves, R. Breaux,

and D. a. Graeber, Usability engineering of virtual
environments (VEs): identifying multiple criteria that drive
effective VE system design, International Journal of
Human-Computer Studies, 58(4), pp. 447481, 2003.

[21]

J. Martin Gutierrez, Estudio y evaluacin de contenidos

didcticos en el desarrollo de las habilidades espaciales en
el mbito de la ingeniera. Doctoral Thesis, Universidad
politecnica de Valencia, 2010.

[1]

L. R. Squire, Memory and Brain. Oxford University Press,

1987.

[2]

I. E. Dror, Technology enhanced learning: The good, the

bad, and the ugly, Pragmatics & Cognition, vol. 16, no. 2,
pp. 215223, Jan. 2008.

[3]

[4]

[5]

[6]

T. L. Andersen, S. Kristensen, B. W. Nielsen, and K.

Grnbk, Designing Augmented Reality Board Games :
The BattleBoard 3D experience, in Proceeding of the
2004 conference on Interaction design and children:
building a community, 2004, pp. 137138.
R. Galantay, J. Torpus, and M. Engeli, living-room.
Interactive, Space-Oriented Augmented Reality,
Proceedings of 12th annual ACM conference on
Multimedia - MULTIMEDIA04, 2004, pp. 6471.
B. Brederode, P. Markopoulos, M. Gielen, A. Vermeeren,
and H. de Ridder, pOwerball: The design of a novel
mixed-reality game for children with mixed abilities,
Proceedings 2005 conference on Interaction design and
children - IDC 05, 2005, pp. 3239.
Z. Pan, A. D. Cheok, H. Yang, J. Zhu, and J. Shi, Virtual
reality and mixed reality for virtual learning
environments, Computers & Graphics, vol. 30, no. 1, pp.
2028, Feb. 2006.

[7]

R. Budhiraja, S. Verma, and A. Pandey, Designing

interactive presentation systems for classrooms,
Proceedings of 28th ACM International Conference on
Design of Communication - SIGDOC10, 2010, p. 259.

[8]

C. Alvarez, R. Alarcon, and M. Nussbaum, Implementing

collaborative learning activities in the classroom supported
by one-to-one mobile computing: A design-based process,
Journal of Systems and Software, vol. 84, no. 11, pp.
19611976, Nov. 2011.

42